2020: Week 26 - Solution




You can find our full solution workflow below and download it here. 


This week the challenge was all about working with unique IDs and how to deal with things when they don't quite all match up!

Step 1 - Join Inputs

The first step this week is to bring in both the 'Internal Data' and '3rd Party Data' tables and then identify which IDs match up perfectly, and the ones that don't. To do this we can use the following joins: 

Perfect Match
This is an inner join on ID = 3rd Party ID


We can then add a step after the join where we remove the [Scent-1] field and add a string calculation called 'Status' that labels all of the rows as 'Matched'.

Unmatched Internal
Right Outer Join to include all values from the Internal table, joining on 3rd Party ID = ID


Similar to before, we are then going to add a clean step, then remove any 3rd Party fields (ID & Sales) and also rename [Scent-1] to [Scent]

Unmatched 3rd Party
Left Outer Join to include all values from the 3rd Party table, joining on 3rd Party ID = ID


Again we will add a clean step, and this time just removing the ID, Scent-1, & Sales fields. 

We should now have three branches to our workflow that looks something like this: 

Step 2 - Prepare Unmatched Fields

Now we have joined our data together, it's now time to focus on the unmatched fields. Therfore, we are going to use another join to bring these together, however this time we will be joining on Scent from both sides: 


Now we have both our Internal and 3rd Party unmatched IDs together, we need to calculate the difference in sales between the two. We can use the following calculation: 

Sales Difference (Absolute Value)
abs([Sales]-[3rd Party Sales])
We now want to calculate the minimum sales difference across our data set for the 3rd Party IDs, therefore we use the following LOD to calculate this: 

Min Sales Difference



We can then use this field to filter our any values that are equal to the sales difference. So the filter calculation is [Min Sales Difference] = [Sales Difference (Absolute Value)] then we can also remove the Min Sales Difference field. 

Next, we want to do the same process but for our Internal IDs so we follow these steps: 

1. Calculate Min Sales Difference for IDs


2. Filter on this condition - [Min Sales Difference (Internal)]=[Sales Difference (Absolute Value)]

3. Remove [Min Sales Difference (Internal)] & [Sales Difference (Absolute Value)]

We have now identified the products that can be matched by their scent, therefore we need a calculation to keep this status. We can use a string like before, to create Status as 'Matched on Scent' for all of the rows. 

The final step of this part of the challenge is to recombine the two workflows that we currently have, so we can use a union tool to stack the 'Perfect Matches' onto top of the matches that we have just identified by Scent. Our table should look something like this: 


Step 3 - Classify Unmatched

Now we have identified the IDs that match from the different tables, it's now time to identify the ones that don't. To do this we are going to bring in the two inputs again, you can use the original ones but bringing in new ones allows us to keep the workflow nice and clean!

We are going to use the two inputs to again join with our original workflow, creating two new streams for Internal and 3rd Party IDs. 

Internal Unmatched 
Outer Right Join including all information from the Internal Data table. Join on ID = ID


We can then clean the table with the following steps: 
  • Remove Status, ID, Scent, Sales, 3rd Party ID, 3rd Party Sales
  • Rename ID-1 to ID, Scent-1 to Scent, Sales-1 to Sales
  • String calculation Status - 'Unmatched - Internal'
As a result of this we should have the following table: 


3rd Party Unmatched
We are going to use a similar process with the 3rd Party unmatched branch, so we will made the following steps: 

1. Join 3rd Party Data to Original workflow. This is an Outer Right Join with all of the 3rd Party rows returned, and joined on 3rd Party ID = 3rd Party ID


2. Clean the following fields: 
  • Remove Fields - Status, ID, Scent, Sales, 3rd Party ID, 3rd Party Sales
  • Rename 3rd Party ID-1 to 3rd Party ID, Scent-1 to Scent, 3rd Party Sales-1 to 3rd Party Sales
  • String calculation for Status - 'Unmatched - 3rd Party' 
Now we have prepared both the Internal and 3rd Party IDs we can bring these together into one table by using the union tool. These tools look like this in our workflow: 

Our table should look something like this: 


The final step is to bring all of this back together with our original workflow. As the data is structured in the same way throughout the workflow, we can take our current workflow (Unmatched Union) and the workflow before the unmatched work, and union these together. 


As a result of this final union, we now have the desired output with the Matched and both Unmatched IDs in one table: 


Make sure to fill in the participation tracker, share using #PreppinData on Twitter and post you solutions onto our Tableau Forums community page so that we can compare our workflows! 

Popular posts from this blog

2023: Week 1 The Data Source Bank

Image
 Created by: Carl Allchin Welcome to a New Year of Preppin' Data. These are weekly exercises to help you learn and develop data preparation skills. We publish the challenges on a Wednesday and share a solution the following Tuesday. You can take the challenges whenever you want and we love to see your solutions. With data preparation, there is never just one way to complete the tasks so sharing your solutions will help others learn too. Share on Twitter, LinkedIn, the Tableau Forums or wherever you want to too. Tag Jenny Martin, Tom Prowse or myself or just use the #PreppinData to share your solutions.  The challenges are designed for learning Tableau Prep but we have a broad community who complete the challenges in R, Python, SQL, DBT, EasyMorph and many other tools. We love seeing people learn new tools so feel free to use whatever tools you want to complete the challenges.  A New Year means we start afresh so January's challenges will be focused on beginners. We will use dif
Read more

2023: Week 2 - International Bank Account Numbers

Image
Challenge By: Jenny Martin For week 2 of our beginner month, Data Source Bank has a requirement to construct International Bank Account Numbers (IBANs), even for Transactions taking place in the UK. We have all the information in separate fields, we just need to put it altogether in the following order: Inputs A list of the transactions, with information about the receiving bank account  A lookup table for the SWIFT Bank Codes  Requirements Input the data In the Transactions table, there is a Sort Code field which contains dashes. We need to remove these so just have a 6 digit string ( hint ) Use the SWIFT Bank Code lookup table to bring in additional information about the SWIFT code and Check Digits of the receiving bank account ( hint ) Add a field for the Country Code ( hint ) Hint: all these transactions take place in the UK so the Country Code should be GB Create the IBAN as above ( hint ) Hint: watch out for trying to combine sting fields with numeric fields - check data types Re
Read more

How to...Handle Free Text

Image
Free isn't always a good thing. In data, Free text is the example to state when proving that statements correct. However, lots of benefit can be gained from understanding data that has been entered in Free Text fields. What do we mean by Free Text? Free Text is the string based data that comes from allowing people to type answers in to systems and forms. The resulting data is normally stored within one column, with one answer per cell. As Free Text means the answer could be anything, this is what you get - absolutely anything. From expletives to slang, the words you will find in the data may be a challenge to interpret but the text is the closest way to collect the voice of your customer / employee. The Free Text field is likely to contain long, rambling sentences that can simply be analysed. If you count these fields, you are likely to have one of each entry each. Therefore, simply counting the entries will not provide anything meaningful to your analysis. The value is in
Read more